Abstract: A method of performing predictive analysis on an external index number describing quality of an information system based on a dynamic causal loop diagram of a plurality of quality meters describing the information system is disclosed. In an aspect, one or more quality meters are selected (402) from the plurality of quality meters for predicting values of an external index number. Values of selected quality meters and the external index number maintained (401) in a database are processed (403) using one or more statistical analysis methods to determine dependencies between the plurality of quality meters and the external index number. Based on the processing, a prediction model for predicting external index number values is established (404). Said model may be trained (406) based on received training data (405) using machine learning tools.

Abstract: A method of performing predictive analysis on an external index number describing quality of an information system based on a dynamic causal loop diagram of a plurality of quality meters describing the information system is disclosed. In an aspect, one or more quality meters are selected (402) from the plurality of quality meters for predicting values of an external index number. Values of selected quality meters and the external index number maintained (401) in a database are processed (403) using one or more statistical analysis methods to determine dependencies between the plurality of quality meters and the external index number. Based on the processing, a prediction model for predicting external index number values is established (404). Said model may be trained (406) based on received training data (405) using machine learning tools.

Abstract: A method, apparatus and system for updating a causal loop diagram of quality meters for an information system, including retrieving the causal loop diagram of the quality meters for the information system, the diagram comprising at least a first quality meter, and a second quality meter and an interaction between them with its strength. The invention further includes receiving online measurement data for the first quality meter; receiving online measurement data for the second quality meter; processing the received online measurement data by statistical analysis to determine a measured interaction with its strength between the first quality meter and the second quality meter; comparing the measured interaction with its strength to the retrieved interaction with its strength; and updating the retrieved interaction in the causal loop diagram of quality meters for an information system, if the measured interaction with its strength is different to the retrieved interaction with its strength.